fault_32.c 16.9 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
/*
 *  linux/arch/i386/mm/fault.c
 *
 *  Copyright (C) 1995  Linus Torvalds
 */

#include <linux/signal.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/types.h>
#include <linux/ptrace.h>
#include <linux/mman.h>
#include <linux/mm.h>
#include <linux/smp.h>
#include <linux/interrupt.h>
#include <linux/init.h>
#include <linux/tty.h>
#include <linux/vt_kern.h>		/* For unblank_screen() */
#include <linux/highmem.h>
22
#include <linux/bootmem.h>		/* for max_low_pfn */
23
#include <linux/vmalloc.h>
L
Linus Torvalds 已提交
24
#include <linux/module.h>
25
#include <linux/kprobes.h>
26
#include <linux/uaccess.h>
27
#include <linux/kdebug.h>
28
#include <linux/kprobes.h>
L
Linus Torvalds 已提交
29 30 31

#include <asm/system.h>
#include <asm/desc.h>
32
#include <asm/segment.h>
L
Linus Torvalds 已提交
33 34 35

extern void die(const char *,struct pt_regs *,long);

36 37
#ifdef CONFIG_KPROBES
static inline int notify_page_fault(struct pt_regs *regs)
38
{
39 40 41 42 43 44 45 46 47
	int ret = 0;

	/* kprobe_running() needs smp_processor_id() */
	if (!user_mode_vm(regs)) {
		preempt_disable();
		if (kprobe_running() && kprobe_fault_handler(regs, 14))
			ret = 1;
		preempt_enable();
	}
48

49
	return ret;
50
}
51 52
#else
static inline int notify_page_fault(struct pt_regs *regs)
53
{
54
	return 0;
55
}
56
#endif
57

L
Linus Torvalds 已提交
58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78
/*
 * Return EIP plus the CS segment base.  The segment limit is also
 * adjusted, clamped to the kernel/user address space (whichever is
 * appropriate), and returned in *eip_limit.
 *
 * The segment is checked, because it might have been changed by another
 * task between the original faulting instruction and here.
 *
 * If CS is no longer a valid code segment, or if EIP is beyond the
 * limit, or if it is a kernel address when CS is not a kernel segment,
 * then the returned value will be greater than *eip_limit.
 * 
 * This is slow, but is very rarely executed.
 */
static inline unsigned long get_segment_eip(struct pt_regs *regs,
					    unsigned long *eip_limit)
{
	unsigned long eip = regs->eip;
	unsigned seg = regs->xcs & 0xffff;
	u32 seg_ar, seg_limit, base, *desc;

79 80 81 82 83 84 85
	/* Unlikely, but must come before segment checks. */
	if (unlikely(regs->eflags & VM_MASK)) {
		base = seg << 4;
		*eip_limit = base + 0xffff;
		return base + (eip & 0xffff);
	}

L
Linus Torvalds 已提交
86
	/* The standard kernel/user address space limit. */
87
	*eip_limit = user_mode(regs) ? USER_DS.seg : KERNEL_DS.seg;
L
Linus Torvalds 已提交
88 89
	
	/* By far the most common cases. */
90
	if (likely(SEGMENT_IS_FLAT_CODE(seg)))
L
Linus Torvalds 已提交
91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107
		return eip;

	/* Check the segment exists, is within the current LDT/GDT size,
	   that kernel/user (ring 0..3) has the appropriate privilege,
	   that it's a code segment, and get the limit. */
	__asm__ ("larl %3,%0; lsll %3,%1"
		 : "=&r" (seg_ar), "=r" (seg_limit) : "0" (0), "rm" (seg));
	if ((~seg_ar & 0x9800) || eip > seg_limit) {
		*eip_limit = 0;
		return 1;	 /* So that returned eip > *eip_limit. */
	}

	/* Get the GDT/LDT descriptor base. 
	   When you look for races in this code remember that
	   LDT and other horrors are only used in user space. */
	if (seg & (1<<2)) {
		/* Must lock the LDT while reading it. */
108
		mutex_lock(&current->mm->context.lock);
L
Linus Torvalds 已提交
109 110 111 112
		desc = current->mm->context.ldt;
		desc = (void *)desc + (seg & ~7);
	} else {
		/* Must disable preemption while reading the GDT. */
113
 		desc = (u32 *)get_cpu_gdt_table(get_cpu());
L
Linus Torvalds 已提交
114 115 116 117 118 119 120
		desc = (void *)desc + (seg & ~7);
	}

	/* Decode the code segment base from the descriptor */
	base = get_desc_base((unsigned long *)desc);

	if (seg & (1<<2)) { 
121
		mutex_unlock(&current->mm->context.lock);
L
Linus Torvalds 已提交
122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139
	} else
		put_cpu();

	/* Adjust EIP and segment limit, and clamp at the kernel limit.
	   It's legitimate for segments to wrap at 0xffffffff. */
	seg_limit += base;
	if (seg_limit < *eip_limit && seg_limit >= base)
		*eip_limit = seg_limit;
	return eip + base;
}

/* 
 * Sometimes AMD Athlon/Opteron CPUs report invalid exceptions on prefetch.
 * Check that here and ignore it.
 */
static int __is_prefetch(struct pt_regs *regs, unsigned long addr)
{ 
	unsigned long limit;
140
	unsigned char *instr = (unsigned char *)get_segment_eip (regs, &limit);
L
Linus Torvalds 已提交
141 142 143 144 145 146 147 148 149
	int scan_more = 1;
	int prefetch = 0; 
	int i;

	for (i = 0; scan_more && i < 15; i++) { 
		unsigned char opcode;
		unsigned char instr_hi;
		unsigned char instr_lo;

150
		if (instr > (unsigned char *)limit)
L
Linus Torvalds 已提交
151
			break;
152
		if (probe_kernel_address(instr, opcode))
L
Linus Torvalds 已提交
153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176
			break; 

		instr_hi = opcode & 0xf0; 
		instr_lo = opcode & 0x0f; 
		instr++;

		switch (instr_hi) { 
		case 0x20:
		case 0x30:
			/* Values 0x26,0x2E,0x36,0x3E are valid x86 prefixes. */
			scan_more = ((instr_lo & 7) == 0x6);
			break;
			
		case 0x60:
			/* 0x64 thru 0x67 are valid prefixes in all modes. */
			scan_more = (instr_lo & 0xC) == 0x4;
			break;		
		case 0xF0:
			/* 0xF0, 0xF2, and 0xF3 are valid prefixes */
			scan_more = !instr_lo || (instr_lo>>1) == 1;
			break;			
		case 0x00:
			/* Prefetch instruction is 0x0F0D or 0x0F18 */
			scan_more = 0;
177
			if (instr > (unsigned char *)limit)
L
Linus Torvalds 已提交
178
				break;
179
			if (probe_kernel_address(instr, opcode))
L
Linus Torvalds 已提交
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
				break;
			prefetch = (instr_lo == 0xF) &&
				(opcode == 0x0D || opcode == 0x18);
			break;			
		default:
			scan_more = 0;
			break;
		} 
	}
	return prefetch;
}

static inline int is_prefetch(struct pt_regs *regs, unsigned long addr,
			      unsigned long error_code)
{
	if (unlikely(boot_cpu_data.x86_vendor == X86_VENDOR_AMD &&
		     boot_cpu_data.x86 >= 6)) {
		/* Catch an obscure case of prefetch inside an NX page. */
		if (nx_enabled && (error_code & 16))
			return 0;
		return __is_prefetch(regs, addr);
	}
	return 0;
} 

205 206 207 208 209 210 211 212 213 214 215 216
static noinline void force_sig_info_fault(int si_signo, int si_code,
	unsigned long address, struct task_struct *tsk)
{
	siginfo_t info;

	info.si_signo = si_signo;
	info.si_errno = 0;
	info.si_code = si_code;
	info.si_addr = (void __user *)address;
	force_sig_info(si_signo, &info, tsk);
}

L
Linus Torvalds 已提交
217 218
fastcall void do_invalid_op(struct pt_regs *, unsigned long);

219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address)
{
	unsigned index = pgd_index(address);
	pgd_t *pgd_k;
	pud_t *pud, *pud_k;
	pmd_t *pmd, *pmd_k;

	pgd += index;
	pgd_k = init_mm.pgd + index;

	if (!pgd_present(*pgd_k))
		return NULL;

	/*
	 * set_pgd(pgd, *pgd_k); here would be useless on PAE
	 * and redundant with the set_pmd() on non-PAE. As would
	 * set_pud.
	 */

	pud = pud_offset(pgd, address);
	pud_k = pud_offset(pgd_k, address);
	if (!pud_present(*pud_k))
		return NULL;

	pmd = pmd_offset(pud, address);
	pmd_k = pmd_offset(pud_k, address);
	if (!pmd_present(*pmd_k))
		return NULL;
247
	if (!pmd_present(*pmd)) {
248
		set_pmd(pmd, *pmd_k);
249 250
		arch_flush_lazy_mmu_mode();
	} else
251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281
		BUG_ON(pmd_page(*pmd) != pmd_page(*pmd_k));
	return pmd_k;
}

/*
 * Handle a fault on the vmalloc or module mapping area
 *
 * This assumes no large pages in there.
 */
static inline int vmalloc_fault(unsigned long address)
{
	unsigned long pgd_paddr;
	pmd_t *pmd_k;
	pte_t *pte_k;
	/*
	 * Synchronize this task's top level page-table
	 * with the 'reference' page table.
	 *
	 * Do _not_ use "current" here. We might be inside
	 * an interrupt in the middle of a task switch..
	 */
	pgd_paddr = read_cr3();
	pmd_k = vmalloc_sync_one(__va(pgd_paddr), address);
	if (!pmd_k)
		return -1;
	pte_k = pte_offset_kernel(pmd_k, address);
	if (!pte_present(*pte_k))
		return -1;
	return 0;
}

282 283
int show_unhandled_signals = 1;

L
Linus Torvalds 已提交
284 285 286 287 288 289 290 291 292
/*
 * This routine handles page faults.  It determines the address,
 * and the problem, and then passes it off to one of the appropriate
 * routines.
 *
 * error_code:
 *	bit 0 == 0 means no page found, 1 means protection fault
 *	bit 1 == 0 means read, 1 means write
 *	bit 2 == 0 means kernel, 1 means user-mode
293 294
 *	bit 3 == 1 means use of reserved bit detected
 *	bit 4 == 1 means fault was an instruction fetch
L
Linus Torvalds 已提交
295
 */
296 297
fastcall void __kprobes do_page_fault(struct pt_regs *regs,
				      unsigned long error_code)
L
Linus Torvalds 已提交
298 299 300 301 302
{
	struct task_struct *tsk;
	struct mm_struct *mm;
	struct vm_area_struct * vma;
	unsigned long address;
303
	int write, si_code;
N
Nick Piggin 已提交
304
	int fault;
L
Linus Torvalds 已提交
305 306

	/* get the address */
307
        address = read_cr2();
L
Linus Torvalds 已提交
308 309 310

	tsk = current;

311
	si_code = SEGV_MAPERR;
L
Linus Torvalds 已提交
312 313 314 315 316 317 318 319 320 321 322 323

	/*
	 * We fault-in kernel-space virtual memory on-demand. The
	 * 'reference' page table is init_mm.pgd.
	 *
	 * NOTE! We MUST NOT take any locks for this case. We may
	 * be in an interrupt or a critical region, and should
	 * only copy the information from the master page table,
	 * nothing more.
	 *
	 * This verifies that the fault happens in kernel space
	 * (error_code & 4) == 0, and that the fault was not a
324
	 * protection error (error_code & 9) == 0.
L
Linus Torvalds 已提交
325
	 */
326 327 328
	if (unlikely(address >= TASK_SIZE)) {
		if (!(error_code & 0x0000000d) && vmalloc_fault(address) >= 0)
			return;
329
		if (notify_page_fault(regs))
330 331
			return;
		/*
L
Linus Torvalds 已提交
332 333 334 335
		 * Don't take the mm semaphore here. If we fixup a prefetch
		 * fault we could otherwise deadlock.
		 */
		goto bad_area_nosemaphore;
336 337
	}

338
	if (notify_page_fault(regs))
339 340 341 342 343 344
		return;

	/* It's safe to allow irq's after cr2 has been saved and the vmalloc
	   fault has been handled. */
	if (regs->eflags & (X86_EFLAGS_IF|VM_MASK))
		local_irq_enable();
L
Linus Torvalds 已提交
345 346 347 348 349 350 351 352 353 354 355 356 357

	mm = tsk->mm;

	/*
	 * If we're in an interrupt, have no user context or are running in an
	 * atomic region then we must not take the fault..
	 */
	if (in_atomic() || !mm)
		goto bad_area_nosemaphore;

	/* When running in the kernel we expect faults to occur only to
	 * addresses in user space.  All other faults represent errors in the
	 * kernel and should generate an OOPS.  Unfortunatly, in the case of an
A
Adrian Bunk 已提交
358
	 * erroneous fault occurring in a code path which already holds mmap_sem
L
Linus Torvalds 已提交
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385
	 * we will deadlock attempting to validate the fault against the
	 * address space.  Luckily the kernel only validly references user
	 * space from well defined areas of code, which are listed in the
	 * exceptions table.
	 *
	 * As the vast majority of faults will be valid we will only perform
	 * the source reference check when there is a possibilty of a deadlock.
	 * Attempt to lock the address space, if we cannot we then validate the
	 * source.  If this is invalid we can skip the address space check,
	 * thus avoiding the deadlock.
	 */
	if (!down_read_trylock(&mm->mmap_sem)) {
		if ((error_code & 4) == 0 &&
		    !search_exception_tables(regs->eip))
			goto bad_area_nosemaphore;
		down_read(&mm->mmap_sem);
	}

	vma = find_vma(mm, address);
	if (!vma)
		goto bad_area;
	if (vma->vm_start <= address)
		goto good_area;
	if (!(vma->vm_flags & VM_GROWSDOWN))
		goto bad_area;
	if (error_code & 4) {
		/*
386 387 388 389
		 * Accessing the stack below %esp is always a bug.
		 * The large cushion allows instructions like enter
		 * and pusha to work.  ("enter $65535,$31" pushes
		 * 32 pointers and then decrements %esp by 65535.)
L
Linus Torvalds 已提交
390
		 */
391
		if (address + 65536 + 32 * sizeof(unsigned long) < regs->esp)
L
Linus Torvalds 已提交
392 393 394 395 396 397 398 399 400
			goto bad_area;
	}
	if (expand_stack(vma, address))
		goto bad_area;
/*
 * Ok, we have a good vm_area for this memory access, so
 * we can handle it..
 */
good_area:
401
	si_code = SEGV_ACCERR;
L
Linus Torvalds 已提交
402 403 404
	write = 0;
	switch (error_code & 3) {
		default:	/* 3: write, present */
405
				/* fall through */
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413
		case 2:		/* write, not present */
			if (!(vma->vm_flags & VM_WRITE))
				goto bad_area;
			write++;
			break;
		case 1:		/* read, present */
			goto bad_area;
		case 0:		/* read, not present */
414
			if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
L
Linus Torvalds 已提交
415 416 417 418 419 420 421 422 423
				goto bad_area;
	}

 survive:
	/*
	 * If for any reason at all we couldn't handle the fault,
	 * make sure we exit gracefully rather than endlessly redo
	 * the fault.
	 */
N
Nick Piggin 已提交
424 425 426
	fault = handle_mm_fault(mm, vma, address, write);
	if (unlikely(fault & VM_FAULT_ERROR)) {
		if (fault & VM_FAULT_OOM)
L
Linus Torvalds 已提交
427
			goto out_of_memory;
N
Nick Piggin 已提交
428 429 430
		else if (fault & VM_FAULT_SIGBUS)
			goto do_sigbus;
		BUG();
L
Linus Torvalds 已提交
431
	}
N
Nick Piggin 已提交
432 433 434 435
	if (fault & VM_FAULT_MAJOR)
		tsk->maj_flt++;
	else
		tsk->min_flt++;
L
Linus Torvalds 已提交
436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457

	/*
	 * Did it hit the DOS screen memory VA from vm86 mode?
	 */
	if (regs->eflags & VM_MASK) {
		unsigned long bit = (address - 0xA0000) >> PAGE_SHIFT;
		if (bit < 32)
			tsk->thread.screen_bitmap |= 1 << bit;
	}
	up_read(&mm->mmap_sem);
	return;

/*
 * Something tried to access memory that isn't in our memory map..
 * Fix it, but check if it's kernel or user first..
 */
bad_area:
	up_read(&mm->mmap_sem);

bad_area_nosemaphore:
	/* User mode accesses just cause a SIGSEGV */
	if (error_code & 4) {
458 459 460 461 462
		/*
		 * It's possible to have interrupts off here.
		 */
		local_irq_enable();

L
Linus Torvalds 已提交
463 464 465 466 467 468 469
		/* 
		 * Valid to do another page fault here because this one came 
		 * from user space.
		 */
		if (is_prefetch(regs, address, error_code))
			return;

470 471 472 473 474 475 476 477
		if (show_unhandled_signals && unhandled_signal(tsk, SIGSEGV) &&
		    printk_ratelimit()) {
			printk("%s%s[%d]: segfault at %08lx eip %08lx "
			    "esp %08lx error %lx\n",
			    tsk->pid > 1 ? KERN_INFO : KERN_EMERG,
			    tsk->comm, tsk->pid, address, regs->eip,
			    regs->esp, error_code);
		}
L
Linus Torvalds 已提交
478 479 480 481
		tsk->thread.cr2 = address;
		/* Kernel addresses are always protection faults */
		tsk->thread.error_code = error_code | (address >= TASK_SIZE);
		tsk->thread.trap_no = 14;
482
		force_sig_info_fault(SIGSEGV, si_code, address, tsk);
L
Linus Torvalds 已提交
483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521
		return;
	}

#ifdef CONFIG_X86_F00F_BUG
	/*
	 * Pentium F0 0F C7 C8 bug workaround.
	 */
	if (boot_cpu_data.f00f_bug) {
		unsigned long nr;
		
		nr = (address - idt_descr.address) >> 3;

		if (nr == 6) {
			do_invalid_op(regs, 0);
			return;
		}
	}
#endif

no_context:
	/* Are we prepared to handle this kernel fault?  */
	if (fixup_exception(regs))
		return;

	/* 
	 * Valid to do another page fault here, because if this fault
	 * had been triggered by is_prefetch fixup_exception would have 
	 * handled it.
	 */
 	if (is_prefetch(regs, address, error_code))
 		return;

/*
 * Oops. The kernel tried to access some bad page. We'll have to
 * terminate things with extreme prejudice.
 */

	bust_spinlocks(1);

522
	if (oops_may_print()) {
523 524 525
		__typeof__(pte_val(__pte(0))) page;

#ifdef CONFIG_X86_PAE
526 527 528 529 530 531 532 533
		if (error_code & 16) {
			pte_t *pte = lookup_address(address);

			if (pte && pte_present(*pte) && !pte_exec_kernel(*pte))
				printk(KERN_CRIT "kernel tried to execute "
					"NX-protected page - exploit attempt? "
					"(uid: %d)\n", current->uid);
		}
534
#endif
535 536 537 538 539 540 541
		if (address < PAGE_SIZE)
			printk(KERN_ALERT "BUG: unable to handle kernel NULL "
					"pointer dereference");
		else
			printk(KERN_ALERT "BUG: unable to handle kernel paging"
					" request");
		printk(" at virtual address %08lx\n",address);
P
Pavel Emelyanov 已提交
542
		printk(KERN_ALERT "printing eip: %08lx ", regs->eip);
543 544 545 546

		page = read_cr3();
		page = ((__typeof__(page) *) __va(page))[address >> PGDIR_SHIFT];
#ifdef CONFIG_X86_PAE
P
Pavel Emelyanov 已提交
547
		printk("*pdpt = %016Lx ", page);
548 549 550 551 552
		if ((page >> PAGE_SHIFT) < max_low_pfn
		    && page & _PAGE_PRESENT) {
			page &= PAGE_MASK;
			page = ((__typeof__(page) *) __va(page))[(address >> PMD_SHIFT)
			                                         & (PTRS_PER_PMD - 1)];
P
Pavel Emelyanov 已提交
553
			printk(KERN_ALERT "*pde = %016Lx ", page);
554 555 556
			page &= ~_PAGE_NX;
		}
#else
P
Pavel Emelyanov 已提交
557
		printk("*pde = %08lx ", page);
L
Linus Torvalds 已提交
558
#endif
559 560 561 562 563 564 565 566 567 568 569 570

		/*
		 * We must not directly access the pte in the highpte
		 * case if the page table is located in highmem.
		 * And let's rather not kmap-atomic the pte, just in case
		 * it's allocated already.
		 */
		if ((page >> PAGE_SHIFT) < max_low_pfn
		    && (page & _PAGE_PRESENT)) {
			page &= PAGE_MASK;
			page = ((__typeof__(page) *) __va(page))[(address >> PAGE_SHIFT)
			                                         & (PTRS_PER_PTE - 1)];
P
Pavel Emelyanov 已提交
571
			printk("*pte = %0*Lx ", sizeof(page)*2, (u64)page);
572
		}
P
Pavel Emelyanov 已提交
573 574

		printk("\n");
575 576
	}

577 578 579
	tsk->thread.cr2 = address;
	tsk->thread.trap_no = 14;
	tsk->thread.error_code = error_code;
L
Linus Torvalds 已提交
580 581 582 583 584 585 586 587 588 589
	die("Oops", regs, error_code);
	bust_spinlocks(0);
	do_exit(SIGKILL);

/*
 * We ran out of memory, or some other thing happened to us that made
 * us unable to handle the page fault gracefully.
 */
out_of_memory:
	up_read(&mm->mmap_sem);
590
	if (is_init(tsk)) {
L
Linus Torvalds 已提交
591 592 593 594 595 596
		yield();
		down_read(&mm->mmap_sem);
		goto survive;
	}
	printk("VM: killing process %s\n", tsk->comm);
	if (error_code & 4)
597
		do_group_exit(SIGKILL);
L
Linus Torvalds 已提交
598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613
	goto no_context;

do_sigbus:
	up_read(&mm->mmap_sem);

	/* Kernel mode? Handle exceptions or die */
	if (!(error_code & 4))
		goto no_context;

	/* User space => ok to do another page fault */
	if (is_prefetch(regs, address, error_code))
		return;

	tsk->thread.cr2 = address;
	tsk->thread.error_code = error_code;
	tsk->thread.trap_no = 14;
614
	force_sig_info_fault(SIGBUS, BUS_ADRERR, address, tsk);
615
}
L
Linus Torvalds 已提交
616

617 618 619 620 621 622 623 624 625 626 627
void vmalloc_sync_all(void)
{
	/*
	 * Note that races in the updates of insync and start aren't
	 * problematic: insync can only get set bits added, and updates to
	 * start are only improving performance (without affecting correctness
	 * if undone).
	 */
	static DECLARE_BITMAP(insync, PTRS_PER_PGD);
	static unsigned long start = TASK_SIZE;
	unsigned long address;
L
Linus Torvalds 已提交
628

629 630 631
	if (SHARED_KERNEL_PMD)
		return;

632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651
	BUILD_BUG_ON(TASK_SIZE & ~PGDIR_MASK);
	for (address = start; address >= TASK_SIZE; address += PGDIR_SIZE) {
		if (!test_bit(pgd_index(address), insync)) {
			unsigned long flags;
			struct page *page;

			spin_lock_irqsave(&pgd_lock, flags);
			for (page = pgd_list; page; page =
					(struct page *)page->index)
				if (!vmalloc_sync_one(page_address(page),
								address)) {
					BUG_ON(page != pgd_list);
					break;
				}
			spin_unlock_irqrestore(&pgd_lock, flags);
			if (!page)
				set_bit(pgd_index(address), insync);
		}
		if (address == start && test_bit(pgd_index(address), insync))
			start = address + PGDIR_SIZE;
L
Linus Torvalds 已提交
652 653
	}
}